Am I oversharpening

Discussion in 'Digital Cameras' started by Celtic Boar, May 30, 2005.

  1. Celtic Boar

    Skip M Guest

    I'd like to purchase some 56mm film, Steve, where can I do so? And I'd
    really like to talk to a member of this "everyone who is someone," too.
    If Bayer type cameras have "crippling optical flaws," why is it you claim to
    have bought and used a Canon 1D mkII?
    Skip M, Jun 14, 2005
    1. Advertisements

  2. SNIP
    Only in your twisted imagination.

    Only in your twisted imagination.
    It may occur if the Anti-aliasing prefilter is not strong enough an
    the Raw converter favors detail above accuracy, but it is not a "huge"

    Wow, it took you a whole sentence to contradict yourself.
    You're wrong, again, as usual.
    First of all there is no such thing as a monochrome sensor. They are
    sensitive to the full human visible spectrum, and then some. Second,
    there is no subdivision. Each pixel has it's own potential well for
    storage of converted photon energy. What you attempt to describe, can
    be an issue with the Foveon design, where the 3 charges from a single
    stacked sampling position have to be stored in a limited capacity.
    You are wrong, again, as usual.
    Foveon sensors filter the three spectral bands as a function of
    penetration depth. Obviously photons can only be converted once, at
    the wavelength dependent depth. Bayer CFA sensor designs do filter out
    some of the light (roughly 2/3 rd of the spectrum per sensel), but not
    necessarily of the incident radiation/energy. That depends on quantum
    efficiency which varies by wavelength.

    What's more, the filtered wavelengths per sensel are reconstructed
    upon demosaicing, so there is no "dimming" of the image. Nor is there
    an "incredible loss of optical data", luminance is sampled at each
    sensor, Anti-aliasing filters provide additional luminocity to
    surrounding sensels, and basically only the lower sampling density of
    color will somewhat reduce color resolution. Since luminance
    resolution is more important to the human eye, that is usually not a
    big problem.

    All this, wrong on at least 5 counts, does show your lack of
    understanding of how sensor designs really work.

    The rest of your delusions are too weird to justify further comments,
    you are simply clueless.

    Bart van der Wolf, Jun 14, 2005
    1. Advertisements

  3. Sure Steve (aka George Preddy), now be a nice 39 year old boy and take
    your medicine.

    Bart van der Wolf, Jun 14, 2005
  4. Celtic Boar

    JPS Guest

    In message <NCbQsJA$>,
    Huh? Any photon that hits a stack can no longer do anything but
    register in on of the layers, or be missed. Blurring would require it
    to bleed into another stack, about as far away as the width of a stack.
    JPS, Jun 14, 2005
  5. No, pixel count isn't meaningless, otherwise you wouldn't lie about it
    all the time.

    upscaling and interpolating are not the same thing. There are 8MP of
    data in a 20D image. It is interpolated to produce a full colour
    image, but there are 8MP of distinct data.
    No a Sigma is 3.43 MP, everything else is upscaling. There is only
    3.43 MP of Data in an SD9/10 image. producing any more MP is only
    increased number of duplicate pixels.

    grant kinsley, Jun 15, 2005
  6. Celtic Boar

    Ken Tough Guest

    Yes, good point. Once it's hits a "well", it's there.
    Ken Tough, Jun 15, 2005
  7. Celtic Boar

    Bill Funk Guest

    So what?
    Sigma claims the SDx are 10.3MP cameras; what do they know?
    Trying to redefine industry-wide standards to meet the desires of
    Sigma/Foveon is your best trick, but it's all smoke. Not even mirrors,
    just smoke.
    Bill Funk, Jun 19, 2005
  8. SNIP
    Scanners don't use Bayer CFA sensors. Most desktop models use
    tri-linear sensor arrays, sometimes in a staggered layout (double
    lines per color but with half a pixel offset).

    Bart van der Wolf, Jun 19, 2005
  9. spread over 3.4 MP of resolution

    .. The 1DMKII only has 8M RGB sensors.
    spread over 8MP of resolution.

    grant kinsley, Jun 21, 2005
  10. 1 guy who's current involvement is......
    But the professional camera community still uses those Bayer sensor
    cameras, I guess reality and Dr. Mead's opinion (if he even said that,
    and knowing the absolute lack of truth in statements, I doubt he did)
    don't jibe.

    I like reality better, you like your own little fantasy world.

    Steve, just how much Sigma stock do you own anyway,

    grant kinsley, Jun 21, 2005
  11. Celtic Boar

    Ken Tough Guest

    Again, wrong. I already pointed out the problem in what you write.
    It is not 10.4MP sensors, since MP stands for "mega pixels". It is
    2268 x 1512 x 3 colour sensors = 10.4M sensors. In terms of full
    colour pixels it is 2268 x 1512 = 3.43 MP.

    I hate to repeat it again, but every time you write it someone has
    to follow up with a correction for the Google history.
    Ken Tough, Jun 21, 2005
  12. Source?

    The man may be trying to push his own product, he may use smoke and
    mirrors while doing it, but even he knows the laws of physics cannot
    be put aside. He is VERY unlikely to have talked about your "the
    scalable interpolation unit is 2x2 sensors wide", because it isn't.
    It's more likely your lack of understanding the fundamentals of
    sampling theory and Bayer CFA reconstruction.
    So you've just insulted the man's intelligence by taking yourself as a
    reference, what's next?

    Bart van der Wolf, Jun 21, 2005
  13. Celtic Boar

    JPS Guest

    In message <>,
    Wrong. The SD9 has 3.43M quasi-RGB sensels. The 1DmkII has no RGB
    sensors, but has 4M green sensors, and 2M each of red and blue.
    JPS, Jun 23, 2005
  14. Celtic Boar

    JPS Guest

    In message <>,
    He is either wrong, or you are misquoting him.
    The panchromatic sensor is not a bayer filter.
    There's bound to be one flop in the mix.
    That doesn't increase his smartness a bit.
    JPS, Jun 23, 2005
  15. Celtic Boar

    JPS Guest

    In message <>,
    A million billion zillion psychological shares.
    JPS, Jun 23, 2005
  16. Celtic Boar

    Ray Fischer Guest

    I agree the SD9 is a 3.43MP camera,
    "George Preddy" <> in <bq6bp8$rep$>
    Ray Fischer, Jun 23, 2005
  17. Celtic Boar

    Skip M Guest

    Could you please provide a link to this quote? All I can find is Phil
    discussing how the X3 sensor is the equivalent of a 6mp Bayer sensor, and
    has some imaging issues like color bleed...
    Skip M, Jun 25, 2005
  18. Oh? A reference? What search terms did you use?

    The CCD was invented decades ago at Bell Labs by Willard Boyle and
    George Smith. No sign of participation by Carver Mead.

    He apparently *did* help invent a CMOS "adaptive photoreceptor"
    circuit, but I didn't see any reference indicating that it is used in
    current CMOS-sensor digital cameras either.

    Carver Mead did a variety of other interesting things, but not what
    you're attributing to him.

    Can you point to any reference that supports your statement, outside of
    Foveon propaganda?

    Dave Martindale, Jun 25, 2005
  19. Celtic Boar

    JPS Guest

    In message <>,
    He (you) said that he invented *Bayer*. That does not follow from
    inventing the panchromatic sensor.
    Of course, that was an initial impression. He has had more time to look
    more closely. What does he say now?
    JPS, Jun 25, 2005
  20. Celtic Boar

    Skip M Guest

    Just what the heck does gender have to do with it?
    Skip M, Jun 25, 2005
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Similar Threads
There are no similar threads yet.